http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49586
--- Comment #4 from Tobias Burnus <burnus at gcc dot gnu.org> 2011-06-30 13:57:21 UTC --- (In reply to comment #3) > Untested fix for the order. For the example of comment 0, gfortran + patch gives the gfortran 4.1 et al. result. For the example which is contained in the attachment, one gets the following result, when printing (in that order) i, j, k, l. gfortran 4.1-4.5: ICE in gfc_assign_data_value_range. gfortran 4.7 (w/ patch) -- and pathf90 3.2.99, openf95 4.2 and sunf95 8.3: 4, 5, 5, 5, 5, 1, 2, 2, 2, 3, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 6 4, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 6 4, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 6 4, 1, 5, 5, 5, 5, 2, 5, 5, 5, 5, 2, 5, 5, 5, 5, 2, 5, 5, 5, 5, 3, 5, 5, 6 ifort 11.1 and pgf90 10.1: 4, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 6 4, 1, 0, 0, 0, 0, 2, 0, 0, 0, 0, 2, 0, 0, 0, 0, 2, 0, 0, 0, 0, 3, 0, 0, 6 4, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 6 4, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 6 gfortran 4.6/4.7 (w/o patch): 4, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 6 4, 1, 5, 5, 5, 5, 2, 5, 5, 5, 5, 2, 5, 5, 5, 5, 2, 5, 5, 5, 5, 3, 5, 5, 6 4, 5, 5, 5, 5, 1, 2, 2, 2, 3, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 6 4, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 6 For comparison, the m and n lines are: 4, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 6 4, 5, 5, 5, 5, 1, 2, 2, 2, 3, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 6
